Previous research has investigated the possibility of facilitating the driver detection of road surface type by means of selective manipulation of the steering-wheel acceleration signal. In previous studies a selective increase in acceleration amplitude has been found to facilitate road-surface-type detection, as has selective manipulation of the individual transient events which are present in the signal. The previous research results have been collected into a first guideline for the optimization of the steering-wheel acceleration signal, and the guideline has been tested in the current study. The test stimuli used in the current study were ten steering-wheel acceleration-time histories which were selected from an extensive database of road test measurements performed by the research group. The time histories, which were all from midsized European automobiles and European roads, were selected such that the widest possible operating envelope could be achieved in terms of the r.m.s. value of the steering acceleration, the kurtosis, the power spectral density function, and the number of transient events present in the signal. The time histories were manipulated by means of the mildly nonstationary mission synthesis algorithm in order to increase, by a factor of 2, both the number and the size of the transient events contained within the frequency interval from 20 Hz to 60 Hz. The ensemble, composed of both the unmanipulated and the manipulated time histories, was used to perform a laboratory-based detection task with 15 participants, who were presented the individual stimuli in random order. The participants were asked to state, by answering 'yes' or 'no', whether each stimulus was considered to be from the road surface that was displayed in front of them by means of a large photograph on a board. The results suggest that the selectively manipulated steering-wheel acceleration stimuli produced improved detection for eight of the ten road surface types which were tested, with a maximum improvement of 14 per cent in the case of the broken road surface. The selective manipulation did lead, however, to some degradation in detection for the motorway road stimulus and for the noise road stimulus, thus suggesting that the current guideline is not universally optimal for all road surfaces.