Photometry of the short-period (P ~0.285 d) low-mass (M₁ ~0.67, M₂ ~0.48 M_⊙) W UMa-type eclipsing binary RW Dor, mainly involving a good series of observations in 1987–88, using standard B and V filters at the Auckland Observatory, is examined. Analysis via light curve fittings alone permits both transit (near main sequence) and occultation (W-type contact binary) solutions. Consideration of proximity or other effects on the relatively small number of measured radial velocities gives rise to a more detailed review of the alternatives, though the (W-type) configuration determined by Hilditch, Hill & Bell (1992) is still confirmed. Further implications of this configuration are examined, given an apparent general trend to period decrease (Δ P/P ~6 × 10^–11), and some light curve asymmetry. This is consistent with the initially low-mass companion, Case-B scenario of Budding (1984). The evidence does suggest irregularity of the period variation however, and further detailed surveillance, including more spectroscopy, is desirable.