Background: COVID-19 infection has poor outcomes for patients (pts) with cancer. Understanding vaccine response as a correlate of protection from severe infection is essential to advise pts regarding protective behaviours and optimal vaccine schedule. This Australian cohort is unique due to low rates of COVID-19 exposure at study entry (July-November 2021). and use of a 3 dose schedule. Pts initially received 2 doses of either BNT162b2 (Pf) at a 3 week interval, or ChadOx1-S (AZ) at a 6 week interval, all then received a 3^rd dose, either mRNA-1273 (Mod) or Pf after 2-4 months, and finally a 4^th dose at an interval of a further 3 months, for a subset. Methods: SerOzNET (ACTRN12621001004853) has enrolled pts with solid and haematological (haem) cancers prior to initial vaccination. Serial blood samples were processed for serum, PBMC and PMN at timepoints: 0, then 3-4 weeks post dose 1 then 2 then 3 then 4 (where administered). We report here neutralizing antibodies (nAb) against wild type (wt) and delta and omicron variants of concern (VOC); quantitative S-protein IgG antibody (Abbott); T-cell correlates measured by levels of interferon-g (IFN g), tumour necrosis factor-a, interleukins (IL-) 2/4/5/13; and epigenetic profiling of T cells. Results: The cohort consists of 401 pts with median age 58 (range 18-85); 59% female; 128 (32%) haem cancers. 377 (94%) are on current or recent (< 12 months) systemic therapy: 162 (43%) chemotherapy, 62 (16%) immunotherapy, 40 (10%) combined chemo/immunotherapy, 113 (29%) hormonal or targeted therapy. 42 (10%) received anti-CD20 therapy < 12 months, 6 (1.4%) had allogeneic stem cell transplant. NAb levels against wt are available for 256 pts post dose 1, 245 pts post dose 2 and 159 pts post dose 3 (will be updated). Response rates post dose were respectively 27%, 77% and 88%. Pts with haem cancer were less likely to respond to vaccination at any time compared to pts with solid cancer (p < 0.001, chi-squared test). After 3 doses, 3.8% of pts with solid cancer and 27.8% with haem cancer lacked NAb. NAb results to VOC delta are available for 92 pts post dose 2: 25/92 (27%) were negative, compared with a non-response rate to wt of 15% at same time in same pts. IFN-γ-Spike response was detectable in 18/31 (58%) and 24/30 (80%) pts post dose 1 and 2 respectively. 101 pts to date have received a 4^th dose; data will be available at the meeting, as will epigenetic profiles and detailed clinicopathological correlations. Conclusions: This interim analysis shows that a significant proportion of pts with haem cancers (27.8%) lack protective Sars-CoV-2 antibodies following 3 vaccinations, whereas only 3.8% of solid cancer pts lack detectable response. Results from other B and T cell parameters may also be important in identifying pts less well protected by vaccination. Follow up is ongoing, response rate post 4^th dose will be presented at the meeting. Clinical trial information: ACTRN12621001004853.