Abstract
Some results are given on the form of the information matrices, canonical efficiency factors and average efficiencies for various classes of row-column designs. All the designs considered are structurally incomplete, that is not all cells of the design are assigned treatments (or treatment combinations), and the designs have column balance, that is every pair of columns in the design intersects rows in exactly f non-empty cells, where f is independent of the choice of columns. An upper bound is given for the average efficiency factor of a row-column design which has adjusted orthogonality and is arranged as a square layout with one empty cell in each row and column.