diff --git a/checkpoint_test.go b/checkpoint_test.go
index 074bb46e8..daa54df19 100644
--- a/checkpoint_test.go
+++ b/checkpoint_test.go
@@ -15,7 +15,6 @@
 package tsdb
 
 import (
-	"fmt"
 	"io/ioutil"
 	"os"
 	"path/filepath"
@@ -82,7 +81,7 @@ func TestDeleteCheckpoints(t *testing.T) {
 func TestCheckpoint(t *testing.T) {
 	dir, err := ioutil.TempDir("", "test_checkpoint")
 	testutil.Ok(t, err)
-	fmt.Println(dir)
+	defer os.RemoveAll(dir)
 
 	var enc RecordEncoder
 	// Create a dummy segment to bump the initial number.
@@ -138,11 +137,10 @@ func TestCheckpoint(t *testing.T) {
 	}
 	testutil.Ok(t, w.Close())
 
-	stats, err = Checkpoint(nil, w, 100, 106, func(x uint64) bool {
+	_, err = Checkpoint(nil, w, 100, 106, func(x uint64) bool {
 		return x%2 == 0
 	}, last/2)
 	testutil.Ok(t, err)
-	testutil.Equals(t, 106, stats.HighSegment)
 
 	// Only the new checkpoint should be left.
 	files, err := fileutil.ReadDir(dir)
diff --git a/db.go b/db.go
index fcfbeeeb2..df17991eb 100644
--- a/db.go
+++ b/db.go
@@ -37,6 +37,7 @@ import (
 	"github.com/prometheus/tsdb/chunkenc"
 	"github.com/prometheus/tsdb/fileutil"
 	"github.com/prometheus/tsdb/labels"
+	"github.com/prometheus/tsdb/wal"
 	"golang.org/x/sync/errgroup"
 )
 
@@ -221,18 +222,18 @@ func Open(dir string, l log.Logger, r prometheus.Registerer, opts *Options) (db
 		return nil, errors.Wrap(err, "create leveled compactor")
 	}
 
-	wal, err := OpenSegmentWAL(filepath.Join(dir, "wal"), l, opts.WALFlushInterval, r)
+	wlog, err := wal.New(l, r, filepath.Join(dir, "wal"))
 	if err != nil {
 		return nil, err
 	}
-	db.head, err = NewHead(r, l, wal, opts.BlockRanges[0])
+	db.head, err = NewHead(r, l, wlog, opts.BlockRanges[0])
 	if err != nil {
 		return nil, err
 	}
 	if err := db.reload(); err != nil {
 		return nil, err
 	}
-	if err := db.head.ReadWAL(); err != nil {
+	if err := db.head.Init(); err != nil {
 		return nil, errors.Wrap(err, "read WAL")
 	}
 
diff --git a/head.go b/head.go
index 372842865..4e12369da 100644
--- a/head.go
+++ b/head.go
@@ -15,6 +15,7 @@ package tsdb
 
 import (
 	"math"
+	"path/filepath"
 	"runtime"
 	"sort"
 	"strings"
@@ -30,6 +31,7 @@ import (
 	"github.com/prometheus/tsdb/chunks"
 	"github.com/prometheus/tsdb/index"
 	"github.com/prometheus/tsdb/labels"
+	"github.com/prometheus/tsdb/wal"
 )
 
 var (
@@ -53,9 +55,10 @@ var (
 type Head struct {
 	chunkRange int64
 	metrics    *headMetrics
-	wal        WAL
+	wal        *wal.WAL
 	logger     log.Logger
 	appendPool sync.Pool
+	bytesPool  sync.Pool
 
 	minTime, maxTime int64
 	lastSeriesID     uint64
@@ -169,13 +172,10 @@ func newHeadMetrics(h *Head, r prometheus.Registerer) *headMetrics {
 }
 
 // NewHead opens the head block in dir.
-func NewHead(r prometheus.Registerer, l log.Logger, wal WAL, chunkRange int64) (*Head, error) {
+func NewHead(r prometheus.Registerer, l log.Logger, wal *wal.WAL, chunkRange int64) (*Head, error) {
 	if l == nil {
 		l = log.NewNopLogger()
 	}
-	if wal == nil {
-		wal = NopWAL()
-	}
 	if chunkRange < 1 {
 		return nil, errors.Errorf("invalid chunk range %d", chunkRange)
 	}
@@ -206,6 +206,8 @@ func (h *Head) processWALSamples(
 ) (unknownRefs uint64) {
 	defer close(output)
 
+	maxt := h.MaxTime()
+
 	for samples := range input {
 		for _, s := range samples {
 			if s.T < mint || s.Ref%total != partition {
@@ -221,17 +223,27 @@ func (h *Head) processWALSamples(
 				h.metrics.chunksCreated.Inc()
 				h.metrics.chunks.Inc()
 			}
+			if s.T > maxt {
+				maxt = s.T
+			}
 		}
 		output <- samples
 	}
+
+	for {
+		ht := h.MaxTime()
+		if maxt <= ht {
+			break
+		}
+		if atomic.CompareAndSwapInt64(&h.maxTime, ht, maxt) {
+			break
+		}
+	}
+
 	return unknownRefs
 }
 
-// ReadWAL initializes the head by consuming the write ahead log.
-func (h *Head) ReadWAL() error {
-	defer h.postings.EnsureOrder()
-
-	r := h.wal.Reader()
+func (h *Head) loadWAL(r *wal.Reader) error {
 	mint := h.MinTime()
 
 	// Track number of samples that referenced a series we don't know about
@@ -263,49 +275,71 @@ func (h *Head) ReadWAL() error {
 		input = output
 	}
 
-	// TODO(fabxc): series entries spread between samples can starve the sample workers.
-	// Even with bufferd channels, this can impact startup time with lots of series churn.
-	// We must not paralellize series creation itself but could make the indexing asynchronous.
-	seriesFunc := func(series []RefSeries) {
-		for _, s := range series {
-			h.getOrCreateWithID(s.Ref, s.Labels.Hash(), s.Labels)
+	var (
+		dec     RecordDecoder
+		series  []RefSeries
+		samples []RefSample
+		tstones []Stone
+	)
+	for r.Next() {
+		series, samples, tstones = series[:0], samples[:0], tstones[:0]
+		rec := r.Record()
 
-			if h.lastSeriesID < s.Ref {
-				h.lastSeriesID = s.Ref
+		switch dec.Type(rec) {
+		case RecordSeries:
+			series, err := dec.Series(rec, series)
+			if err != nil {
+				return errors.Wrap(err, "decode series")
 			}
-		}
-	}
-	samplesFunc := func(samples []RefSample) {
-		// We split up the samples into chunks of 5000 samples or less.
-		// With O(300 * #cores) in-flight sample batches, large scrapes could otherwise
-		// cause thousands of very large in flight buffers occupying large amounts
-		// of unused memory.
-		for len(samples) > 0 {
-			n := 5000
-			if len(samples) < n {
-				n = len(samples)
-			}
-			var buf []RefSample
-			select {
-			case buf = <-input:
-			default:
-			}
-			firstInput <- append(buf[:0], samples[:n]...)
-			samples = samples[n:]
-		}
-	}
-	deletesFunc := func(stones []Stone) {
-		for _, s := range stones {
-			for _, itv := range s.intervals {
-				if itv.Maxt < mint {
-					continue
+			for _, s := range series {
+				h.getOrCreateWithID(s.Ref, s.Labels.Hash(), s.Labels)
+
+				if h.lastSeriesID < s.Ref {
+					h.lastSeriesID = s.Ref
 				}
-				h.tombstones.addInterval(s.ref, itv)
 			}
+		case RecordSamples:
+			samples, err := dec.Samples(rec, samples)
+			if err != nil {
+				return errors.Wrap(err, "decode samples")
+			}
+			// We split up the samples into chunks of 5000 samples or less.
+			// With O(300 * #cores) in-flight sample batches, large scrapes could otherwise
+			// cause thousands of very large in flight buffers occupying large amounts
+			// of unused memory.
+			for len(samples) > 0 {
+				n := 5000
+				if len(samples) < n {
+					n = len(samples)
+				}
+				var buf []RefSample
+				select {
+				case buf = <-input:
+				default:
+				}
+				firstInput <- append(buf[:0], samples[:n]...)
+				samples = samples[n:]
+			}
+		case RecordTombstones:
+			tstones, err := dec.Tombstones(rec, tstones)
+			if err != nil {
+				return errors.Wrap(err, "decode tombstones")
+			}
+			for _, s := range tstones {
+				for _, itv := range s.intervals {
+					if itv.Maxt < mint {
+						continue
+					}
+					h.tombstones.addInterval(s.ref, itv)
+				}
+			}
+		default:
+			return errors.Errorf("invalid record type %v", dec.Type(rec))
 		}
 	}
-
-	err := r.Read(seriesFunc, samplesFunc, deletesFunc)
+	if r.Err() != nil {
+		return errors.Wrap(r.Err(), "read records")
+	}
 
 	// Signal termination to first worker and wait for last one to close its output channel.
 	close(firstInput)
@@ -313,16 +347,58 @@ func (h *Head) ReadWAL() error {
 	}
 	wg.Wait()
 
-	if err != nil {
-		return errors.Wrap(err, "consume WAL")
-	}
 	if unknownRefs > 0 {
-		level.Warn(h.logger).Log("msg", "unknown series references in WAL samples", "count", unknownRefs)
+		level.Warn(h.logger).Log("msg", "unknown series references", "count", unknownRefs)
 	}
 	return nil
 }
 
-// Truncate removes all data before mint from the head block and truncates its WAL.
+// Init loads data from the write ahead log and prepares the head for writes.
+func (h *Head) Init() error {
+	defer h.postings.EnsureOrder()
+
+	if h.wal == nil {
+		return nil
+	}
+
+	// Backfill the checkpoint first if it exists.
+	cp, n, err := LastCheckpoint(h.wal.Dir())
+	if err != nil && err != ErrNotFound {
+		return errors.Wrap(err, "find last checkpoint")
+	}
+	if err == nil {
+		sr, err := wal.NewSegmentsReader(filepath.Join(h.wal.Dir(), cp))
+		if err != nil {
+			return errors.Wrap(err, "open checkpoint")
+		}
+		defer sr.Close()
+
+		// A corrupted checkpoint is a hard error for now and requires user
+		// intervention. There's likely little data that can be recovered anyway.
+		if err := h.loadWAL(wal.NewReader(sr)); err != nil {
+			return errors.Wrap(err, "backfill checkpoint")
+		}
+		n++
+	}
+
+	// Backfill segments from the last checkpoint onwards
+	sr, err := wal.NewSegmentsRangeReader(h.wal.Dir(), n, -1)
+	if err != nil {
+		return errors.Wrap(err, "open WAL segments")
+	}
+	defer sr.Close()
+
+	err = h.loadWAL(wal.NewReader(sr))
+	if err == nil {
+		return nil
+	}
+	if err := h.wal.Repair(err); err != nil {
+		return errors.Wrap(err, "repair corrupted WAL")
+	}
+	return nil
+}
+
+// Truncate removes old data before mint from the head.
 func (h *Head) Truncate(mint int64) error {
 	initialize := h.MinTime() == math.MinInt64
 
@@ -348,18 +424,37 @@ func (h *Head) Truncate(mint int64) error {
 	level.Info(h.logger).Log("msg", "head GC completed", "duration", time.Since(start))
 	h.metrics.gcDuration.Observe(time.Since(start).Seconds())
 
+	if h.wal == nil {
+		return nil
+	}
 	start = time.Now()
 
+	m, n, err := h.wal.Segments()
+	if err != nil {
+		return errors.Wrap(err, "get segment range")
+	}
+	n-- // Never consider last segment for checkpoint.
+	if n < 0 {
+		return nil // no segments yet.
+	}
+	// The lower third of segments should contain mostly obsolete samples.
+	// If we have too few segments, it's not worth checkpointing yet.
+	n = m + (n-m)/3
+	if n <= m {
+		return nil
+	}
+
 	keep := func(id uint64) bool {
 		return h.series.getByID(id) != nil
 	}
-	if err := h.wal.Truncate(mint, keep); err == nil {
-		level.Info(h.logger).Log("msg", "WAL truncation completed", "duration", time.Since(start))
-	} else {
-		level.Error(h.logger).Log("msg", "WAL truncation failed", "err", err, "duration", time.Since(start))
+	if _, err = Checkpoint(h.logger, h.wal, m, n, keep, mint); err != nil {
+		return errors.Wrap(err, "create checkpoint")
 	}
 	h.metrics.walTruncateDuration.Observe(time.Since(start).Seconds())
 
+	level.Info(h.logger).Log("msg", "WAL checkpoint complete",
+		"low", m, "high", n, "duration", time.Since(start))
+
 	return nil
 }
 
@@ -468,6 +563,18 @@ func (h *Head) putAppendBuffer(b []RefSample) {
 	h.appendPool.Put(b[:0])
 }
 
+func (h *Head) getBytesBuffer() []byte {
+	b := h.bytesPool.Get()
+	if b == nil {
+		return make([]byte, 0, 1024)
+	}
+	return b.([]byte)
+}
+
+func (h *Head) putBytesBuffer(b []byte) {
+	h.bytesPool.Put(b[:0])
+}
+
 type headAppender struct {
 	head       *Head
 	mint, maxt int64
@@ -520,15 +627,42 @@ func (a *headAppender) AddFast(ref uint64, t int64, v float64) error {
 	return nil
 }
 
+func (a *headAppender) log() error {
+	if a.head.wal == nil {
+		return nil
+	}
+
+	buf := a.head.getBytesBuffer()
+	defer func() { a.head.putBytesBuffer(buf) }()
+
+	var rec []byte
+	var enc RecordEncoder
+
+	if len(a.series) > 0 {
+		rec = enc.Series(a.series, buf)
+		buf = rec[:0]
+
+		if err := a.head.wal.Log(rec); err != nil {
+			return errors.Wrap(err, "log series")
+		}
+	}
+	if len(a.samples) > 0 {
+		rec = enc.Samples(a.samples, buf)
+		buf = rec[:0]
+
+		if err := a.head.wal.Log(rec); err != nil {
+			return errors.Wrap(err, "log samples")
+		}
+	}
+	return nil
+}
+
 func (a *headAppender) Commit() error {
 	defer a.head.metrics.activeAppenders.Dec()
 	defer a.head.putAppendBuffer(a.samples)
 
-	if err := a.head.wal.LogSeries(a.series); err != nil {
-		return err
-	}
-	if err := a.head.wal.LogSamples(a.samples); err != nil {
-		return errors.Wrap(err, "WAL log samples")
+	if err := a.log(); err != nil {
+		return errors.Wrap(err, "write to WAL")
 	}
 
 	total := len(a.samples)
@@ -568,7 +702,8 @@ func (a *headAppender) Rollback() error {
 
 	// Series are created in the head memory regardless of rollback. Thus we have
 	// to log them to the WAL in any case.
-	return a.head.wal.LogSeries(a.series)
+	a.samples = nil
+	return a.log()
 }
 
 // Delete all samples in the range of [mint, maxt] for series that satisfy the given
@@ -601,8 +736,12 @@ func (h *Head) Delete(mint, maxt int64, ms ...labels.Matcher) error {
 	if p.Err() != nil {
 		return p.Err()
 	}
-	if err := h.wal.LogDeletes(stones); err != nil {
-		return err
+	var enc RecordEncoder
+
+	if h.wal != nil {
+		if err := h.wal.Log(enc.Tombstones(stones, nil)); err != nil {
+			return err
+		}
 	}
 	for _, s := range stones {
 		h.tombstones.addInterval(s.ref, s.intervals[0])
@@ -694,6 +833,9 @@ func (h *Head) MaxTime() int64 {
 
 // Close flushes the WAL and closes the head.
 func (h *Head) Close() error {
+	if h.wal == nil {
+		return nil
+	}
 	return h.wal.Close()
 }
 
diff --git a/head_test.go b/head_test.go
index 9a8c89364..b06a66c26 100644
--- a/head_test.go
+++ b/head_test.go
@@ -14,7 +14,9 @@
 package tsdb
 
 import (
+	"io/ioutil"
 	"math/rand"
+	"os"
 	"testing"
 
 	"github.com/prometheus/tsdb/chunkenc"
@@ -22,6 +24,7 @@ import (
 	"github.com/prometheus/tsdb/index"
 	"github.com/prometheus/tsdb/labels"
 	"github.com/prometheus/tsdb/testutil"
+	"github.com/prometheus/tsdb/wal"
 )
 
 func BenchmarkCreateSeries(b *testing.B) {
@@ -42,42 +45,50 @@ func BenchmarkCreateSeries(b *testing.B) {
 	}
 }
 
-type memoryWAL struct {
-	nopWAL
-	entries []interface{}
-}
-
-func (w *memoryWAL) LogSeries(s []RefSeries) error {
-	w.entries = append(w.entries, s)
-	return nil
-}
-
-func (w *memoryWAL) LogSamples(s []RefSample) error {
-	w.entries = append(w.entries, s)
-	return nil
-}
-
-func (w *memoryWAL) LogDeletes(s []Stone) error {
-	w.entries = append(w.entries, s)
-	return nil
-}
-
-func (w *memoryWAL) Reader() WALReader {
-	return w
-}
-
-func (w *memoryWAL) Read(series func([]RefSeries), samples func([]RefSample), deletes func([]Stone)) error {
-	for _, e := range w.entries {
-		switch v := e.(type) {
+func populateTestWAL(t testing.TB, w *wal.WAL, recs []interface{}) {
+	var enc RecordEncoder
+	for _, r := range recs {
+		switch v := r.(type) {
 		case []RefSeries:
-			series(v)
+			testutil.Ok(t, w.Log(enc.Series(v, nil)))
 		case []RefSample:
-			samples(v)
+			testutil.Ok(t, w.Log(enc.Samples(v, nil)))
 		case []Stone:
-			deletes(v)
+			testutil.Ok(t, w.Log(enc.Tombstones(v, nil)))
 		}
 	}
-	return nil
+}
+
+func readTestWAL(t testing.TB, dir string) (recs []interface{}) {
+	sr, err := wal.NewSegmentsReader(dir)
+	testutil.Ok(t, err)
+	defer sr.Close()
+
+	var dec RecordDecoder
+	r := wal.NewReader(sr)
+
+	for r.Next() {
+		rec := r.Record()
+
+		switch dec.Type(rec) {
+		case RecordSeries:
+			series, err := dec.Series(rec, nil)
+			testutil.Ok(t, err)
+			recs = append(recs, series)
+		case RecordSamples:
+			samples, err := dec.Samples(rec, nil)
+			testutil.Ok(t, err)
+			recs = append(recs, samples)
+		case RecordTombstones:
+			tstones, err := dec.Tombstones(rec, nil)
+			testutil.Ok(t, err)
+			recs = append(recs, tstones)
+		default:
+			t.Fatalf("unknown record type")
+		}
+	}
+	testutil.Ok(t, r.Err())
+	return recs
 }
 
 func TestHead_ReadWAL(t *testing.T) {
@@ -100,13 +111,19 @@ func TestHead_ReadWAL(t *testing.T) {
 			{Ref: 50, T: 101, V: 6},
 		},
 	}
-	wal := &memoryWAL{entries: entries}
+	dir, err := ioutil.TempDir("", "test_read_wal")
+	testutil.Ok(t, err)
+	defer os.RemoveAll(dir)
 
-	head, err := NewHead(nil, nil, wal, 1000)
+	w, err := wal.New(nil, nil, dir)
+	testutil.Ok(t, err)
+	populateTestWAL(t, w, entries)
+
+	head, err := NewHead(nil, nil, w, 1000)
 	testutil.Ok(t, err)
 	defer head.Close()
 
-	testutil.Ok(t, head.ReadWAL())
+	testutil.Ok(t, head.Init())
 	testutil.Equals(t, uint64(100), head.lastSeriesID)
 
 	s10 := head.series.getByID(10)
@@ -259,13 +276,19 @@ func TestHeadDeleteSeriesWithoutSamples(t *testing.T) {
 			{Ref: 50, T: 90, V: 1},
 		},
 	}
-	wal := &memoryWAL{entries: entries}
+	dir, err := ioutil.TempDir("", "test_delete_series")
+	testutil.Ok(t, err)
+	defer os.RemoveAll(dir)
 
-	head, err := NewHead(nil, nil, wal, 1000)
+	w, err := wal.New(nil, nil, dir)
+	testutil.Ok(t, err)
+	populateTestWAL(t, w, entries)
+
+	head, err := NewHead(nil, nil, w, 1000)
 	testutil.Ok(t, err)
 	defer head.Close()
 
-	testutil.Ok(t, head.ReadWAL())
+	testutil.Ok(t, head.Init())
 
 	testutil.Ok(t, head.Delete(0, 100, labels.NewEqualMatcher("a", "1")))
 }
@@ -705,7 +728,7 @@ func TestMemSeries_append(t *testing.T) {
 
 func TestGCChunkAccess(t *testing.T) {
 	// Put a chunk, select it. GC it and then access it.
-	h, err := NewHead(nil, nil, NopWAL(), 1000)
+	h, err := NewHead(nil, nil, nil, 1000)
 	testutil.Ok(t, err)
 	defer h.Close()
 
@@ -745,7 +768,7 @@ func TestGCChunkAccess(t *testing.T) {
 
 func TestGCSeriesAccess(t *testing.T) {
 	// Put a series, select it. GC it and then access it.
-	h, err := NewHead(nil, nil, NopWAL(), 1000)
+	h, err := NewHead(nil, nil, nil, 1000)
 	testutil.Ok(t, err)
 	defer h.Close()
 
@@ -786,7 +809,12 @@ func TestGCSeriesAccess(t *testing.T) {
 }
 
 func TestHead_LogRollback(t *testing.T) {
-	w := &memoryWAL{}
+	dir, err := ioutil.TempDir("", "wal_rollback")
+	testutil.Ok(t, err)
+	defer os.RemoveAll(dir)
+
+	w, err := wal.New(nil, nil, dir)
+	testutil.Ok(t, err)
 	h, err := NewHead(nil, nil, w, 1000)
 	testutil.Ok(t, err)
 
@@ -795,9 +823,11 @@ func TestHead_LogRollback(t *testing.T) {
 	testutil.Ok(t, err)
 
 	testutil.Ok(t, app.Rollback())
-	testutil.Equals(t, 1, len(w.entries))
+	recs := readTestWAL(t, w.Dir())
 
-	series, ok := w.entries[0].([]RefSeries)
-	testutil.Assert(t, ok, "expected series record but got %+v", w.entries[0])
+	testutil.Equals(t, 1, len(recs))
+
+	series, ok := recs[0].([]RefSeries)
+	testutil.Assert(t, ok, "expected series record but got %+v", recs[0])
 	testutil.Equals(t, series, []RefSeries{{Ref: 1, Labels: labels.FromStrings("a", "b")}})
 }