In developing countries, building maintenance is gaining an increasing recognition in many field
of study. Meanwhile, in Nigeria there is lack of effective building maintenance setup that can
sustain the current inadequate housing provision. With ever increasing demand and cost for
housing provision in the country, a lack of appropriate predictive tool for the maintenance of the
existing buildings can have a significant detrimental effect on future housing development.
Presently, there is a remarkable gap and paradigm shift in building maintenance management
from preventive approach to predictive approach that is achievable through building of an
evaluative model to assess a variety of alternative decisions. This paper thus aimed at developing
mathematical models for the maintenance of residential buildings with reference to Niger state,
Nigeria. It identified factors that influence the level of maintenance of residential buildings
standards. The paper employed descriptive and inferential survey research method and subjected
data collected to bi-variate and multi-variate analysis, using Statistical Package for Social
Sciences (SPSS). The analysis enabled the development of a mathematical model which was
tested and found adequate with 94% predictive value respectively for the assessment and
maintenance of residential buildings. The result shows that long term sustainability of these
building could be enhanced through the adoption of appropriate predictive tool for their
maintenance. This will help to achieve their optimum performance throughout their life span.
The paper recommended the adoption of the model for appropriate monitoring and quick
evaluation of residential buildings’ performance in developing countries. It concluded that this
can aid government policy formulation on minimum building maintenance standards to avoid
cumulative decay and deterioration that leads buildings to the point of collapse. It will also assist
architects to make predictions through scientific means rather than perception in their design.