diff --git a/line_benchmark_test.go b/line_benchmark_test.go
index 2711756..80630c2 100644
--- a/line_benchmark_test.go
+++ b/line_benchmark_test.go
@@ -21,8 +21,9 @@ import (
 	"github.com/prometheus/statsd_exporter/pkg/line"
 )
 
-func benchmarkLineToEvents(times int, b *testing.B) {
-	input := []string{
+var (
+	// just a grab bag of mixed formats, valid, invalid
+	mixedLines = []string{
 		"foo1:2|c",
 		"foo2:3|g",
 		"foo3:200|ms",
@@ -35,10 +36,23 @@ func benchmarkLineToEvents(times int, b *testing.B) {
 		"foo15:200|ms:300|ms:5|c|@0.1:6|g\nfoo15a:1|c:5|ms",
 		"some_very_useful_metrics_with_quite_a_log_name:13|c",
 	}
-	logger := log.NewNopLogger()
 
-	// reset benchmark timer to not measure startup costs
-	b.ResetTimer()
+	// The format specific lines have only one line each so the benchmark accurately reflects the time taken to process one line
+	statsdLine           = "foo1:2|c"
+	statsdInvalidLine    = "foo1:2|c||"
+	dogStatsdLine        = "foo1:100|c|#tag1:bar,tag2:baz"
+	dogStatsdInvalidLine = "foo3:100|c|#09digits:0,tag.with.dots:1"
+	signalFxLine         = "foo1.[foo=bar1,dim=val1]test:1|g"
+	signalFxInvalidLine  = "foo1.[foo=bar1,dim=val1test:1|g"
+	influxDbLine         = "foo1,tag1=bar,tag2=baz:100|c"
+	influxDbInvalidLine  = "foo3,tag1=bar,tag2:100|c"
+
+	logger = log.NewNopLogger()
+)
+
+func benchmarkLinesToEvents(times int, b *testing.B, input []string) {
+	// always report allocations since this is a hot path
+	b.ReportAllocs()
 
 	for n := 0; n < b.N; n++ {
 		for i := 0; i < times; i++ {
@@ -49,12 +63,51 @@ func benchmarkLineToEvents(times int, b *testing.B) {
 	}
 }
 
-func BenchmarkLineToEvents1(b *testing.B) {
-	benchmarkLineToEvents(1, b)
+func benchmarkLineToEvents(b *testing.B, inputLine string) {
+	// always report allocations since this is a hot path
+	b.ReportAllocs()
+
+	for n := 0; n < b.N; n++ {
+		line.LineToEvents(inputLine, *sampleErrors, samplesReceived, tagErrors, tagsReceived, logger)
+	}
 }
-func BenchmarkLineToEvents5(b *testing.B) {
-	benchmarkLineToEvents(5, b)
+
+// Mixed statsd formats
+func BenchmarkLineToEventsMixed1(b *testing.B) {
+	benchmarkLinesToEvents(1, b, mixedLines)
 }
-func BenchmarkLineToEvents50(b *testing.B) {
-	benchmarkLineToEvents(50, b)
+func BenchmarkLineToEventsMixed5(b *testing.B) {
+	benchmarkLinesToEvents(5, b, mixedLines)
+}
+func BenchmarkLineToEventsMixed50(b *testing.B) {
+	benchmarkLinesToEvents(50, b, mixedLines)
+}
+
+// Individual format benchmarks
+// Valid Lines
+func BenchmarkLineToEventsStatsd(b *testing.B) {
+	benchmarkLineToEvents(b, statsdLine)
+}
+func BenchmarkLineToEventsDogStatsd(b *testing.B) {
+	benchmarkLineToEvents(b, dogStatsdLine)
+}
+func BenchmarkLineToEventsSignalFx(b *testing.B) {
+	benchmarkLineToEvents(b, signalFxLine)
+}
+func BenchmarkLineToEventsInfluxDb(b *testing.B) {
+	benchmarkLineToEvents(b, influxDbLine)
+}
+
+// Invalid lines
+func BenchmarkLineToEventsStatsdInvalid(b *testing.B) {
+	benchmarkLineToEvents(b, statsdInvalidLine)
+}
+func BenchmarkLineToEventsDogStatsdInvalid(b *testing.B) {
+	benchmarkLineToEvents(b, dogStatsdInvalidLine)
+}
+func BenchmarkLineToEventsSignalFxInvalid(b *testing.B) {
+	benchmarkLineToEvents(b, signalFxInvalidLine)
+}
+func BenchmarkLineToEventsInfluxDbInvalid(b *testing.B) {
+	benchmarkLineToEvents(b, influxDbInvalidLine)
 }