Objective: According to a prior study, there are substantial lapses in journal coverage in the Social Work Abstracts (SWA) database. The current study provides a replication and extension. Method: The longitudinal pattern of coverage of thirty-three journals categorized in SWA as core journals (published in the 1989-1996 period) is examined. Results: The proportion of issues missing from SWA is significantly greater than 0, increase over time, and is significantly biased in favor of NASW journals. Conclusion: The errors in SWA reported here, combined with those previously reported, will exert a long-term negative impact on scholarship unless their existence becomes known to all SWA users and they take steps to compensate for the situation.