There is a paucity of real-world data on vaccine elicited neutralising antibody responses for AZD1222, in African populations following vaccination scale up. Here, we first measured baseline SARS-CoV-2 seroprevalence and levels of protective neutralizing antibodies prior to vaccination rollout using both flow cytometric based analysis of binding antibodies coupled with pseudotyped virus neutralisation assays in two study cohorts from West Africa: Nigerian healthcare workers; (n = 140) and a Ghanaian general community cohort (n = 527). We found that 44% and 28% of pre-vaccination participants showed IgG anti-N positivity, increasing to 59% and 42% respectively with anti-receptor binding dominan (RBD) IgG specific antibodies. The increased prevalence of prior exposure using anti RBD antibodies was corroborated by Pseudotyped virus (PV) neutralizing antibody assays, indicating that overall, 50% of prior infections were missed by N antibody testing. PV titres (serum dilution required to inhibit 50% of infection, ID50) against wild type following 2-dose vaccination regimen were [145 (4.5) to 2579 (4.2) vs 57 (3.0) to 1049 (6.7)] (GMT ± s.d), delta [75 (3.0) to 549 (3.7) vs 37 (2.4) to 453 (7.4)] (GMT ± s.d) and omicron variants [37 (2.4) to 453 (7.4) vs 29 (1.8) to 95 (5.3)] (GMT ± s.d) in the Nigerian (1 month) vs Ghanaian participants (2 months) post vaccination (total n = 94). Previous IgG anti-N was associated with significantly higher neutralizing antibody levels with an observed 3.5-fold [1423 (3.9) vs 4674 (3.4)] (GMT ± s.d) and 2.7-fold [779 (7.1) vs 2128 (4.8) (GMT ± s.d) difference between N positive and negative participants in the Nigerian and Ghanaian cohorts respectively. We also observed serological evidence from N, S and RBD antibodies of breakthrough infection in 8/49 (16%) of Nigerian vaccinees over only 2 months, with neutralisation profiles suggesting delta variant infection consistent with the sampling period when this variant was known to dominate. Importantly, neutralising antibodies waned at 3 months after second dose vs 1 month post second-dose 1695 (4.3)] vs 2579 (4.2) in the Nigerian population who were N negative throughout. IgG anti-N was also observed to wane below cut-off in a total 19/94 (20%) of subjects highlighting the need for a combination of additional markers to characterise previous infection. We conclude that AZD1222 is immunogenic in two independent real world West African cohorts with high background seroprevalence and incidence of breakthrough infection in 2021. Waning titres at 6 months post second dose indicates the need for booster dosing after AZD1222 in the African setting despite hybrid immunity from previous infection.