Our optimised WMISH protocol is not improved by more antibody, PVA or hydrolysed riboprobes. Larvae three to four days post first cleavage (dpfc) were subjected to our optimised WMISH protocol (A, F, K and P). Using a beta tubulin probe, we investigated the effect of increasing the amount of anti-DIG antibody (B, G, L and Q), the addition of PVA to the colour detection solution (C, H, M and R) and the combination of more antibody and PVA (D and I). We also assessed the effect of hydrolysing the engrailed riboprobe individually (N and S) and in combination with a higher antibody concentration and the use of PVA (O and T). None of these modifications generated superior WMISH results to our baseline protocol. Samples incubated in more antibody and developed with PVA showed slightly more intense signals, but a lower signal to noise ratio (B, C, G, H, L, M, Q and R). PVA also appeared to compromise the morphological integrity of older larvae (H, I and R). Signals generated by the hydrolysed engrailed probe were much fainter and were partially masked by an increase in general background staining (N, O, S and T). The optimal treatment (A, F, K, and P) is indicated by a black star. Control WMISH experiments lacking a riboprobe and using the increased antibody concentration do not reveal any staining (E and J). All images of individual larvae are lateral views.