This study proposes an urban Bikeability Index (BI) to assess and prioritise bicycle infrastructure investments, and in turn, improve accessibility for cyclists. The contribution of this paper is twofold. First, we construct a BI that addresses particularities of the roads in cities of the Global South and considers the preferences, perceptions and socioeconomic characteristics by type of cyclist. Second, a tool is developed for decision-makers to prioritise investments in bicycle infrastructure, considering the estimated BI and expected cyclist flows. The methodology is structured into two stages. The first stage constructs a new BI following three steps. First, we select the most important factors and components related to biking behaviour. Second, we estimate weights for selected factors and components by using a discrete choice model to analyse users’ perceptions based on results from ranking surveys. Third, we estimate the BI by type of cyclist as a weighted additive function, which are equivalent to the probabilities calculated using the discrete choice model. The second stage proposes a direct demand model to prioritise bicycle infrastructure investments based on the estimation of BIs, and expected flows of cyclists. We applied the methodology to Barranquilla, Colombia and ranked different bicycle infrastructure investments for increasing bicycle demand in the city. The results suggest that even though primary roads are currently associated with low BI values due to high rates of road accidents, they tend to be preferred by cyclists. The presence of adequate bicycle infrastructure is the most crucial factor for people who frequently cycle for sport, while traffic safety and security are the most critical factors for those who cycle to work.

这项研究提出了城市可骑性指数（BI），以评估和优先考虑自行车基础设施的投资，进而改善骑车者的可及性。本文的贡献是双重的。首先，我们构建一个BI，以解决全球南方城市道路的特殊性，并根据骑自行车者的类型来考虑其偏好，看法和社会经济特征。其次，为决策者开发了一种工具，可以根据估算的BI和预期的骑车人流量对自行车基础设施进行优先投资。该方法分为两个阶段。第一阶段按照以下三个步骤构建新的BI。首先，我们选择与骑车行为有关的最重要因素和组成部分。第二，我们使用离散选择模型根据排名调查的结果来分析用户的看法，从而估算选定因素和组成部分的权重。第三，我们根据骑车人的类型将BI估算为加权加法函数，这与使用离散选择模型计算的概率相等。第二阶段提出了一个直接需求模型，该模型基于BI的估计以及骑自行车者的预期流量来确定自行车基础设施投资的优先级。我们将这种方法应用于哥伦比亚的Barranquilla，并对增加自行车需求的不同自行车基础设施投资进行了排名。结果表明，即使由于道路交通事故的发生率较高，目前主要道路的BI值较低，但骑自行车的人还是倾向于选择主要道路。