To add a bit of how-to detail to
@jtalden's response.
Starting with a typical target curve, which is text file with a series of frequency values with level offsets, like this Dirac target curve with bass boost:
NAME
Harman 6dB
DEVICENAME
Unnamed
BREAKPOINTS
30.00000 6.50000
40.00000 6.36800
50.00000 6.18700
60.00000 5.88900
70.00000 5.46800
80.00000 4.94500
90.00000 4.36600
100.00000 3.78200
110.00000 3.23300
120.00000 2.74600
140.00000 1.98700
160.00000 1.48200
180.00000 1.15900
200.00000 0.95200
220.00000 0.81700
240.00000 0.72800
300.00000 0.59700
400.00000 0.53200
4990.00000 0.0000
20000.00000 -2.32
In REW, click File/Import/Import Frequency Response, and select the target text file:
The result looks like this:
Note that the scale of the imported curve is lower than your REW measurements. Use REW's "Measurement Actions" tool to align the target curve with your speaker measurement levels. You will typically need to add 80-90 dB in the SPL Offset box. You can add or subtract values several times to fine-tune the target curve height.
Here is an example of the final result: