Objectives: To explore whether community health worker household-based maternal health visits improve antenatal care and skilled birth attendance among hard-to-reach fishing villages on Lake Victoria, Uganda. Methods: This quasi-experimental 18-month prospective study involved 486 consenting women aged 15–49 years, who were pregnant or had a pregnancy outcome in the past 6 months, from 6 island fishing communities. The community health worker household-based intervention (community health workers’ household visits to provide counseling, blood pressure measurement, anemia, and HIV testing) involved 243 women from three fishing communities. Random effects logistic regression was used to determine the association between the community health worker intervention and antenatal care and skilled birth attendance among women who had at least 5 months of pregnancy or childbirth at follow-up. Results: Almost all women accepted the community health worker intervention (90.9% (221/243)). Hypertension was at 12.5% (27/216) among those who accepted blood pressure measurements, a third (33.3% (9/27)) were pregnant. HIV prevalence was 23.5% (52/221). Over a third (34.2% (69/202)) of women tested had anemia (hemoglobin levels less than 11 g/dL). The community health worker intervention was associated with attendance of first antenatal care visit within 20 weeks of pregnancy (adjusted odd ratio = 2.1 (95% confidence interval 0.6–7.6)), attendance of at least four antenatal care visits (adjusted odd ratio = 0.9 (95% confidence interval 0.4–2.0)), and skilled birth attendance (adjusted odd ratio = 0.5 (95% confidence interval 0.1–1.5)), though not statistically significant. Conclusion: Community health workers have a crucial role in improving early antenatal care attendance, early community-based diagnosis of anemia, hypertensive disorders, and HIV among women in these hard-to-reach fishing communities.