diff --git a/ber.go b/ber.go
index 53fe04e..71b3c3a 100644
--- a/ber.go
+++ b/ber.go
@@ -9,6 +9,8 @@ import (
 	"math"
 	"os"
 	"reflect"
+	"strconv"
+	"strings"
 	"time"
 	"unicode/utf8"
 )
@@ -391,6 +393,10 @@ func readPacket(reader io.Reader) (*Packet, int, error) {
 			p.Value = DecodeString(content)
 		case TagNULL:
 		case TagObjectIdentifier:
+			oid, err := parseObjectIdentifier(content)
+			if err == nil {
+				p.Value = OIDToString(oid)
+			}
 		case TagObjectDescriptor:
 		case TagExternal:
 		case TagRealFloat:
@@ -406,6 +412,10 @@ func readPacket(reader io.Reader) (*Packet, int, error) {
 				p.Value = val
 			}
 		case TagRelativeOID:
+			oid, err := parseObjectIdentifier(content)
+			if err == nil {
+				p.Value = OIDToString(oid)
+			}
 		case TagSequence:
 		case TagSet:
 		case TagNumericString:
@@ -633,3 +643,166 @@ func NewReal(classType Class, tagType Type, tag Tag, value interface{}, descript
 	}
 	return p
 }
+
+func NewOID(classType Class, tagType Type, tag Tag, value interface{}, description string) *Packet {
+	p := Encode(classType, tagType, tag, nil, description)
+
+	switch v := value.(type) {
+	case string:
+		encoded, err := encodeOID(v)
+		if err != nil {
+			fmt.Printf("failed writing %v", err)
+			return nil
+		}
+		p.Value = v
+		p.Data.Write(encoded)
+		// TODO: support []int already ?
+	default:
+		panic(fmt.Sprintf("Invalid type %T, expected float{64|32}", v))
+	}
+	return p
+}
+
+// encodeOID takes a string representation of an OID and returns its DER-encoded byte slice along with any error.
+func encodeOID(oidString string) ([]byte, error) {
+	// Convert the string representation to an asn1.ObjectIdentifier
+	parts := strings.Split(oidString, ".")
+	oid := make([]int, len(parts))
+	for i, part := range parts {
+		var val int
+		if _, err := fmt.Sscanf(part, "%d", &val); err != nil {
+			return nil, fmt.Errorf("invalid OID part '%s': %w", part, err)
+		}
+		oid[i] = val
+	}
+	if len(oid) < 2 || oid[0] > 2 || (oid[0] < 2 && oid[1] >= 40) {
+		panic(fmt.Sprintf("invalid object identifier % d", oid)) // TODO: not elegant
+	}
+	encoded := make([]byte, 0)
+
+	encoded = appendBase128Int(encoded[:0], int64(oid[0]*40+oid[1]))
+	for i := 2; i < len(oid); i++ {
+		encoded = appendBase128Int(encoded, int64(oid[i]))
+	}
+
+	return encoded, nil
+}
+
+func appendBase128Int(dst []byte, n int64) []byte {
+	l := base128IntLength(n)
+
+	for i := l - 1; i >= 0; i-- {
+		o := byte(n >> uint(i*7))
+		o &= 0x7f
+		if i != 0 {
+			o |= 0x80
+		}
+
+		dst = append(dst, o)
+	}
+
+	return dst
+}
+func base128IntLength(n int64) int {
+	if n == 0 {
+		return 1
+	}
+
+	l := 0
+	for i := n; i > 0; i >>= 7 {
+		l++
+	}
+
+	return l
+}
+
+func OIDToString(oi []int) string {
+	var s strings.Builder
+	s.Grow(32)
+
+	buf := make([]byte, 0, 19)
+	for i, v := range oi {
+		if i > 0 {
+			s.WriteByte('.')
+		}
+		s.Write(strconv.AppendInt(buf, int64(v), 10))
+	}
+
+	return s.String()
+}
+
+// parseObjectIdentifier parses an OBJECT IDENTIFIER from the given bytes and
+// returns it. An object identifier is a sequence of variable length integers
+// that are assigned in a hierarchy.
+func parseObjectIdentifier(bytes []byte) (s []int, err error) {
+	if len(bytes) == 0 {
+		err = fmt.Errorf("zero length OBJECT IDENTIFIER")
+		return
+	}
+
+	// In the worst case, we get two elements from the first byte (which is
+	// encoded differently) and then every varint is a single byte long.
+	s = make([]int, len(bytes)+1)
+
+	// The first varint is 40*value1 + value2:
+	// According to this packing, value1 can take the values 0, 1 and 2 only.
+	// When value1 = 0 or value1 = 1, then value2 is <= 39. When value1 = 2,
+	// then there are no restrictions on value2.
+	v, offset, err := parseBase128Int(bytes, 0)
+	if err != nil {
+		return
+	}
+	if v < 80 {
+		s[0] = v / 40
+		s[1] = v % 40
+	} else {
+		s[0] = 2
+		s[1] = v - 80
+	}
+
+	i := 2
+	for ; offset < len(bytes); i++ {
+		v, offset, err = parseBase128Int(bytes, offset)
+		if err != nil {
+			return
+		}
+		s[i] = v
+	}
+	s = s[0:i]
+	return
+}
+
+// parseBase128Int parses a base-128 encoded int from the given offset in the
+// given byte slice. It returns the value and the new offset.
+func parseBase128Int(bytes []byte, initOffset int) (ret, offset int, err error) {
+	offset = initOffset
+	var ret64 int64
+	for shifted := 0; offset < len(bytes); shifted++ {
+		// 5 * 7 bits per byte == 35 bits of data
+		// Thus the representation is either non-minimal or too large for an int32
+		if shifted == 5 {
+			err = fmt.Errorf("base 128 integer too large")
+			return
+		}
+		ret64 <<= 7
+		b := bytes[offset]
+		// integers should be minimally encoded, so the leading octet should
+		// never be 0x80
+		if shifted == 0 && b == 0x80 {
+			err = fmt.Errorf("integer is not minimally encoded")
+			return
+		}
+		ret64 |= int64(b & 0x7f)
+		offset++
+		if b&0x80 == 0 {
+			ret = int(ret64)
+			// Ensure that the returned value fits in an int on all platforms
+			if ret64 > math.MaxInt32 {
+				err = fmt.Errorf("base 128 integer too large")
+			}
+			return
+		}
+	}
+	err = fmt.Errorf("truncated base 128 integer")
+	return
+}
diff --git a/ber_test.go b/ber_test.go
index 7218c0e..7f0f4ed 100644
--- a/ber_test.go
+++ b/ber_test.go
@@ -100,6 +100,24 @@ func TestString(t *testing.T) {
 	}
 }
 
+func TestEncodeDecodeOID(t *testing.T) {
+	for _, v := range []string{"0.1", "1.1", "2.3", "0.4", "0.4.5.1888", "0.10.5.1888.234.324234"} {
+		enc, err := encodeOID(v)
+		if err != nil {
+			t.Errorf("error on encoding object identifier when encoding %s: %v", v, err)
+		}
+		parsed, err := parseObjectIdentifier(enc)
+		if err != nil {
+			t.Errorf("error on parsing object identifier when parsing %s: %v", v, err)
+		}
+		t.Log(enc)
+		t.Log(OIDToString(parsed))
+		if v != OIDToString(parsed) {
+			t.Error("encoded object identifier did not match parsed")
+		}
+	}
+}
+
 func TestSequenceAndAppendChild(t *testing.T) {
 	values := []string{
 		"HIC SVNT LEONES",