I'm adding this answer to address some of the comments on the other answers that were asking for studies based on small children and/or pregnant women.
First of all, Dr. Rauscher's work on the Mozart Effect has received a lot of bad press in the academic literature:
The successful performance of the Mozart group may be explained by the incomplete use of random assignment of subjects to groups and by experimenter effects in the construction of groups. The results of Rauscher et al. (1998) do not provide strong support for the existence of the Mozart effect.
It doesn't look like the effect exists for young children, either:
We investigated the Mozart effect, as documented by Rauscher, Shaw, and Ky (1993), with school-aged children. Experiment 1 contrasted the spatial IQ scores of children who had listened to a Mozart sonata (K.448) with the scores of children who had listened to a piece of popular dance music in a pretest-post-test design. There was no significant main effect of music and no significant difference between the pretest and post-test scores for both groups. Owing to the non-significant findings, a second experiment was carried out. We used a methodology that had previously replicated the Mozart effect. Again, Expt. 2 did not support the claim that Mozart's music can enhance spatial performance. Groups performed similarly on the control test and the experimental test, irrespective of whether they listened to Mozart or to popular dance music. Since the two different designs produced similar findings, the data suggest that the Mozart effect is so ephemeral that it is questionable as to whether any practical application will come from it. In the discussion, we suggest more fruitful avenues for future research on the relationship between music and spatial performance: arousal and transfer of learning.
[Emphasis is mine.]
But that doesn't really answer the original question: Is classical music good for a pregnant woman and her baby?
To that end, I haven't been able to find any peer reviewed articles that have tested that specific claim. It should be noted, though, that Rauscher—the original discoverer of this "phenomena"—has since admitted that there is no scientific support for the fact that listening to classical music is good for infants:
To date, the Mozart Effect has failed to be replicated in scientific settings on at least a few dozen occasions. Even Rauscher, although she stands by her findings, has been amazed by the appropriation of her work for corporate ends. In a 1999 television debate, Rauscher
agreed with Steele, saying, “There’s no scientific data suggesting that playing Mozart to
babies is going to make them ‘smarter.’”
[Once again, emphasis is mine.]
I will add that this all depends on one's definition of "good". If by "good for a baby" we mean "will increase the baby's intelligence," then I think the answer is: No, listening to classical music probably won't permanently increase a baby's intelligence. It might still be good, though, if it is soothing to the pregnant mother. If the music reduces the mother's stress, this is definitely good since stressors are directly linked to complications in pregnancy. That could be true of any style of music, though.