diff --git a/bench_test.go b/bench_test.go
index 39fdb7c..de25331 100644
--- a/bench_test.go
+++ b/bench_test.go
@@ -54,6 +54,7 @@ func BenchmarkSendRecv(b *testing.B) {
 	recvBuf := make([]byte, 512)
 
 	doneCh := make(chan struct{})
+	b.ResetTimer()
 	go func() {
 		defer close(doneCh)
 		defer server.Close()
diff --git a/util.go b/util.go
index 177eb98..aeccd24 100644
--- a/util.go
+++ b/util.go
@@ -68,15 +68,21 @@ type segmentedBuffer struct {
 	cap uint32
 	len uint32
 	bm  sync.Mutex
-	// read position in b[0].
+	// read position in b[bPos].
 	// We must not reslice any of the buffers in b, as we need to put them back into the pool.
 	readPos int
-	b       [][]byte
+	// bPos is an index in b slice. If bPos == len(b), it means that buffer is empty.
+	bPos int
+	// b is used as a growable buffer. Each Append adds []byte to the end of b.
+	// If there is no space available at the end of the buffer (len(b) == cap(b)), but it has space
+	// at the beginning (bPos > 0 and at least 1/4 of the buffer is empty), data inside b is shifted to the beginning.
+	// Each Read reads from b[bPos] and increments bPos if b[bPos] was fully read.
+	b [][]byte
 }
 
 // NewSegmentedBuffer allocates a ring buffer.
 func newSegmentedBuffer(initialCapacity uint32) segmentedBuffer {
-	return segmentedBuffer{cap: initialCapacity, b: make([][]byte, 0)}
+	return segmentedBuffer{cap: initialCapacity, b: make([][]byte, 0, 16)}
 }
 
 // Len is the amount of data in the receive buffer.
@@ -109,15 +115,15 @@ func (s *segmentedBuffer) GrowTo(max uint32, force bool) (bool, uint32) {
 func (s *segmentedBuffer) Read(b []byte) (int, error) {
 	s.bm.Lock()
 	defer s.bm.Unlock()
-	if len(s.b) == 0 {
+	if s.bPos == len(s.b) {
 		return 0, io.EOF
 	}
-	data := s.b[0][s.readPos:]
+	data := s.b[s.bPos][s.readPos:]
 	n := copy(b, data)
 	if n == len(data) {
-		pool.Put(s.b[0])
-		s.b[0] = nil
-		s.b = s.b[1:]
+		pool.Put(s.b[s.bPos])
+		s.b[s.bPos] = nil
+		s.bPos++
 		s.readPos = 0
 	} else {
 		s.readPos += n
@@ -152,6 +158,23 @@ func (s *segmentedBuffer) Append(input io.Reader, length uint32) error {
 	if n > 0 {
 		s.len += uint32(n)
 		s.cap -= uint32(n)
+		// s.b has no available space at the end, but has space at the beginning
+		if len(s.b) == cap(s.b) && s.bPos > 0 {
+			if s.bPos == len(s.b) {
+				// the buffer is empty, so just move pos
+				s.bPos = 0
+				s.b = s.b[:0]
+			} else if s.bPos > cap(s.b)/4 {
+				// at least 1/4 of buffer is empty, so shift data to the left to free space at the end
+				copied := copy(s.b, s.b[s.bPos:])
+				// clear references to copied data
+				for i := copied; i < len(s.b); i++ {
+					s.b[i] = nil
+				}
+				s.b = s.b[:copied]
+				s.bPos = 0
+			}
+		}
 		s.b = append(s.b, dst[0:n])
 	}
 	return err